Ack is a grep-like tool, optimised for programmers

It&#x27;s also worth checking out The Silver Searcher: <a href="https://github.com/ggreer/the_silver_searcher" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;ggreer&#x2F;the_silver_searcher</a>

I use ag[2], which is pretty much the same as ack, but even faster. The other day I was using it to find all instances in all projects of a list of problematic method names[1], in case anyone wants to see a real world use case.<p>[1]: <a href="http://unix.stackexchange.com/questions/108471/no-output-using-parallel-in-tandem-with-ag-or-ack/108472#108472" rel="nofollow">http:&#x2F;&#x2F;unix.stackexchange.com&#x2F;questions&#x2F;108471&#x2F;no-output-usi...</a><p>[2]: <a href="https://github.com/ggreer/the_silver_searcher" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;ggreer&#x2F;the_silver_searcher</a>

For Java programmers who use Silver Searcher or ack, this lets you search all jars in a directory tree for a given string. Requires GNU Parallel:<p><pre><code> function ffjar() { jars=(.&#x2F;**&#x2F;*.jar) print &quot;Searching ${#jars[*]} jars for &#x27;${*}&#x27;...&quot; parallel --no-notice --tag unzip -l ::: ${jars} | ag ${*} | awk &#x27;{print $1, &quot;:&quot;, $5}&#x27; } </code></pre> Because it uses parallel it spreads the workload across CPUs. I use this frequently when I have to update&#x2F;rewrite&#x2F;create build scripts, and I know a class exists but not which jar file it lives in.

It should be noted that &quot;faster&quot; is due to the fact that it limits itself to searching the subset of files that have code extensions.<p>I&#x27;ve tested grep against ack and ag for large text files and grep won handily, especially the latest version of grep.<p>Also note you can use GNU Parallel to run multiple greps

For some reason, it took me one or two minutes of rereading to realize it was ack, not awk. I think this website was going to be some ironic trash-talking about grep. Then I saw &quot;written in Perl&quot; and I got so confused my head almost exploded.<p>Anyway, neat tool. Will check it out soon.

I&#x27;m assuming the title got rewritten? It is probably important to note:<p>&quot;ack versions 2.00 to 2.11_02 are susceptible to a code execution exploit. Please upgrade to 2.12 or higher ASAP.&quot;

I couldn&#x27;t live without this tool today, very useful to quickly find where <i>that</i> method is used for example.

<a href="https://github.com/monochromegane/the_platinum_searcher" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;monochromegane&#x2F;the_platinum_searcher</a> I use pt for stuff which is not in git, written in go and it is fast

It&#x27;s worth noting that if you&#x27;re always searching in a git repo, git-grep is pretty comparable to ack and you already have it installed.

My ex-colleague introduced this to me and I thank him every time I use ack. It is really so much better than grep. I have set up a bunch of aliases to search by file type and it makes me so productive.

I used Ack a lot when I was coding Perl (it&#x27;s been a while). After I switched to Ruby, I used rak [1], which seemed easier to use most of the time, and nearly identical.<p>However, when you just want to find stuff fast, it&#x27;s annoying to have to deal with Perl&#x2F;CPAN or RVM&#x2F;Rubygems, especially when the dependencies are not installed on your server&#x2F;workstation.<p>That&#x27;s why I&#x27;ve switched to silver searcher (ag) [2], as it can be installed with any OS package manager (brew, apt, yum).<p>[1] <a href="http://rak.rubyforge.org" rel="nofollow">http:&#x2F;&#x2F;rak.rubyforge.org</a> [2] <a href="https://github.com/ggreer/the_silver_searcher" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;ggreer&#x2F;the_silver_searcher</a>

The problem with such tools is often their lack of ubiquity. I don&#x27;t want to start using ack, forget a lot of my grep knowledge, only to ssh into a server and need grep.<p>The benefit of grep&#x27;s ubiquity outweighs any small advantage ack has in usability.

Old thread here: <a href="https://news.ycombinator.com/item?id=6075083" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=6075083</a>

Similar tool in pure Python: <a href="https://github.com/eliben/pss" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;eliben&#x2F;pss</a>

There&#x27;s a list of other tools for search source code besides ack at <a href="http://beyondgrep.com/more-tools/" rel="nofollow">http:&#x2F;&#x2F;beyondgrep.com&#x2F;more-tools&#x2F;</a>, including other grepalikes and indexing tools like ctags and cscope.<p>I suggest that you need not limit yourself to only one tool for your code searching. Toolboxes FTW.

Grep is just as good, and with the recent order of magnitude speed improvement on non-C locales that they made -- see <a href="https://lwn.net/Articles/586899/" rel="nofollow">https:&#x2F;&#x2F;lwn.net&#x2F;Articles&#x2F;586899&#x2F;</a> -- which may not have made its way into distros yet, it&#x27;s easily the best option.<p>I have a simple wrapper over egrep (see <a href="https://github.com/sitaramc/ew" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;sitaramc&#x2F;ew</a> ) that adds those little extras (ignoring binary files, ignoring VCS directories...).<p>I&#x27;m sure it&#x27;s improved since the days I tried it, but I tend to be permanently prejudiced against tools where the author can&#x27;t&#x2F;won&#x27;t document the file selection logic and says &quot;there&#x27;s really no English that explains how it works&quot; when someone asks.

Ack is great, but watch out if you have any source files with unusual file name extensions. Ack will only search file types it knows about. Also if you have your whole source tree in your editor or IDE, then you may as well search there instead.

Yeah, I think grep sucks as well, that&#x27;s why I created my own little ack thing<p>(since it&#x27;s in a sorry state I won&#x27;t post it here, and it will attract the rage of people for not being compatible with grep&#x2F;in python&#x2F;no docs&#x2F;etc)

How does this compare to cscope[0] ?<p>edit: or ctags[1] ?<p>---<p>[0] <a href="http://en.wikipedia.org/wiki/Cscope" rel="nofollow">http:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Cscope</a><p>[1] <a href="http://en.wikipedia.org/wiki/Ctags" rel="nofollow">http:&#x2F;&#x2F;en.wikipedia.org&#x2F;wiki&#x2F;Ctags</a>

I recently found out about git-grep. It&#x27;s good and quite fast.

Is there a tool that is to find what ack is to grep?

I used this for a while, but got bitten by the fact that, by default, it does not search all file types. :-&#x2F;

Been using ack-grep for years. In systems where it&#x27;s not installed I just use grep -Hnir

Is there a code repository somewhere? Can&#x27;t seem to find one...

I&#x27;ll stick with grep, thanks. M-x rgrep in Emacs works great.

This has found it&#x27;s way to HN a long long time ago:<p><a href="https://news.ycombinator.com/item?id=975511" rel="nofollow">https:&#x2F;&#x2F;news.ycombinator.com&#x2F;item?id=975511</a>

Why did they choose the same name as minix compiler ack (<a href="http://tack.sourceforge.net/" rel="nofollow">http:&#x2F;&#x2F;tack.sourceforge.net&#x2F;</a>) ?

And just to warp things even further, use ack to find the files you&#x27;re after, then run through grep:<p><pre><code> ack -f --css | parallel grep search_term</code></pre>

I thought we were done with this already? <a href="https://github.com/ggreer/the_silver_searcher" rel="nofollow">https:&#x2F;&#x2F;github.com&#x2F;ggreer&#x2F;the_silver_searcher</a>

Seen <a href="http://beyondgrep.com/security/" rel="nofollow">http:&#x2F;&#x2F;beyondgrep.com&#x2F;security&#x2F;</a><p>Didn&#x27;t read further.